基于信息融合的医院审计预警模型的研究

发布时间：2018-03-05 12:28

本文选题：信息融合　切入点：审计预警模型　出处：《浙江理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着医院业务的不断扩展和完善,医院审计工作越来越繁杂,传统的审计方式已经无法满足当今数字化、信息化的发展要求,因此连续审计工作将成为今后的发展趋势。虽然信息融合技术在民事领域应用相对广泛,但大多集中在医疗诊断和财务危机方面,在连续审计以及审计预警方面的应用还处在基础研究阶段。本文针对当前医院审计工作的现状,以神经网络BP算法和D-S证据理论为研究核心,构建了基于信息融合的医院审计预警模型,完成审计数据的判断和审计危机预警等功能,从而达到实时监测医院审计活动的目的。本文的主要研究内容概括如下:(1)根据国内外的研究现状,分析并总结了信息融合技术应用于审计预警领域时存在的问题,确定了论文研究的主要内容,并对预警模型的主要理论及技术进行了介绍,为最终系统的设计与实现提供理论依据。(2)针对当前医院审计流程存在的问题,分析审计数据的主要属性并对比审计预警算法的优劣,选取神经网络BP算法构建审计预警模型,同时针对神经网络BP算法所存在的缺陷,提出反向遗传算法(OBGA)与固定隐含层神经元个数相结合的优化方法。通过反向遗传算法找寻输入变量最优组合,隐含层神经元个数提高预测精度的方式,建立OBGA-FHBP审计预警模型。实验对比证明,OBGA-FHBP审计预警模型在准确率、建模时间、网络误差精度以及稳定性上都有所改善。(3)利用审计活动产生的异常数据,采用D-S证据理论的置信函数建立审计危机预警指标优先级排列框架,获得发现异常数据最多的审计危机预警指标,逐步建立DR-Z审计危机预警模型。通过实验证明,该审计危机预警模型能够对医院的经营状况进行正确的预警,并对危机险情做出更加客观的判断,实现事前或事中预警的功能。(4)根据医院审计预警系统的设计目标,对系统进行了总体设计和详细设计,并完成了系统的开发。通过对系统的展示,验证了系统不仅能够对审计数据进行管理与判断,还能够有效地实现医院审计危机预警与风险评估的功能。该系统包括审计数据审核、异常数据处理、危机险情警示以及审计结果公示四个功能模块,实现审计过程的智能化和一体化。
[Abstract]:With the continuous expansion and improvement of hospital business, hospital audit work is becoming more and more complicated, the traditional audit methods can not meet the needs of digital and information development. Therefore, continuous audit work will become a trend in the future. Although information fusion technology is widely used in the civil field, it is mostly focused on medical diagnosis and financial crisis. The application of continuous audit and audit warning is still in the basic research stage. According to the current situation of hospital audit, this paper takes BP algorithm of neural network and D-S evidence theory as the core of the research. A hospital audit early warning model based on information fusion is constructed to complete the functions of judging audit data and early warning of audit crisis. The main research contents of this paper are summarized as follows: (1) according to the current research situation at home and abroad, the problems existing in the application of information fusion technology in the field of audit early warning are analyzed and summarized. The main contents of the paper are determined, and the main theory and technology of the early warning model are introduced, which provides the theoretical basis for the design and implementation of the final system. This paper analyzes the main attributes of audit data and compares the advantages and disadvantages of the audit early warning algorithm, selects the neural network BP algorithm to construct the audit early warning model, and aims at the defects of the neural network BP algorithm. In this paper, an optimization method combining inverse genetic algorithm (OBGA) with the number of neurons in the fixed hidden layer is proposed. The optimal combination of input variables is found through the reverse genetic algorithm, and the number of neurons in the hidden layer improves the prediction accuracy. The experimental results show that the accuracy, modeling time, network error accuracy and stability of OBGA-FHBP audit early warning model have been improved. By using the confidence function of D-S evidence theory, the priority ranking framework of audit crisis warning index is established, and the audit crisis warning index with the most abnormal data is obtained, and the DR-Z audit crisis warning model is established step by step. The early warning model of audit crisis can give a correct early warning to the operating condition of the hospital, and make a more objective judgment on the crisis risk situation, so as to realize the function of early warning in advance or in the event) according to the design goal of the early warning system of the hospital audit, The overall design and detailed design of the system are carried out, and the development of the system is completed. Through the display of the system, it is verified that the system can not only manage and judge the audit data. The system includes four modules: audit data audit, abnormal data processing, crisis warning and audit result announcement. Realize the intelligence and integration of audit process.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R197.3;TP183;TP277

【相似文献】